Dynamics in Behavioral Adaptation to a Transportation Innovation: A Case Study of Carlink–A Smart...

8/15/2019 Dynamics in Behavioral Adaptation to a Transportation Innovation: A Case Study of Carlink–A Smart Carsharing Sy…

1/472

eScholarship provides open access, scholarly publishing

services to the University of California and delivers a dynamic

research platform to scholars worldwide.

Institute of Transportation Studies

UC Davis

Title:

Dynamics in Behavioral Adaptation to a Transportation Innovation: A Case Study of Carlink–ASmart Carsharing System

Author:

Shaheen, Susan, University of California, Berkeley

Publication Date:

09-02-2004

Series:

Recent Work

Permalink:

http://escholarship.org/uc/item/87n6958hAbstract:

Most trips in U.S. metropolitan regions are drive-alone car trips, an expensive and inefficientmeans of moving people. A more efficient system would allow drivers to share cars. Such asystem is often less convenient for travelers, but convenience can be enhanced by deploying“smart” technologies in concert with shared-use vehicles and transit. The motivation for thisresearch is to determine how the use of information and communication technologies can enhanceflexibility and mobility—and what value travelers will place on these new transportation means.My dissertation, using new survey research methods, examines CarLink, a smart carsharingservice designed and deployed under my direction. This dissertation integrates social marketingand learning theories with human activity analysis approaches to explain the processes by whichtravelers can and might accept a transportation innovation. I focus on methods of presentation andlearning to examine response dynamics. To explain the CarLink system to consumers, I developed

several informational media: a brochure, video, and “trial” clinic. My dissertation is based on alongitudinal survey of responses to informational media that I conducted with San Francisco Bay Area residents in the summer of 1998. The survey results provide the attitudinal and belief dataneeded to evaluate dynamics in an individual’s learning and valuing response to an innovation. Toassist in evaluation and interpretation, I also conducted four focus groups, which I moderated, inOctober 1998. I found that willingness to use CarLink was influenced by the amount and type of exposure, as predicted by social marketing and learning theories. Informational media were usedto teach targeted groups, and behavioral modeling (e.g., the video and drive clinic) was introducedto develop participants’ confidence in adopting new behaviors. For instance, participants who onlyread the brochure lost interest over time, while a large majority of those who read the brochure,watched the video, and participated in the clinic, stated that they would use CarLink. I documentedthe process by which individuals moved through definable stages in the behavioral adoptionmodel, from precontemplation to contemplation, and in many cases into action.

Copyright Information: All rights reserved unless otherwise indicated. Contact the author or original publisher for anynecessary permissions. eScholarship is not the copyright owner for deposited works. Learn moreat http://www.escholarship.org/help_copyright.html#reuse

http://escholarship.org/http://www.escholarship.org/help_copyright.html#reusehttp://escholarship.org/uc/item/87n6958hhttp://escholarship.org/uc/itsdavis_rwhttp://escholarship.org/uc/search?creator=Shaheen%2C%20Susanhttp://www.escholarship.org/help_copyright.html#reusehttp://escholarship.org/uc/item/87n6958hhttp://escholarship.org/uc/itsdavis_rwhttp://escholarship.org/uc/search?creator=Shaheen%2C%20Susanhttp://escholarship.org/uc/ucdhttp://escholarship.org/uc/itsdavis_rwhttp://escholarship.org/uc/itsdavis_rwhttp://escholarship.org/http://escholarship.org/http://escholarship.org/http://escholarship.org/


2/472

ISSN 1055-1425

December 1999

This work was performed as part of the California PATH Program of the

University of California, in cooperation with the State of California Business,Transportation, and Housing Agency, Department of Transportation; and the

United States Department of Transportation, Federal Highway Administration.

The contents of this report reflect the views of the authors who are responsible

for the facts and the accuracy of the data presented herein. The contents do not

necessarily reflect the official views or policies of the State of California. This

report does not constitute a standard, specification, or regulation.

Report for MOU 349

CALIFORNIA PATH PROGRAM

INSTITUTE OF TRANSPORTATION STUDIES

UNIVERSITY OF CALIFORNIA, BERKELEY

Dynamics in Behavioral Adaptation to a

Transportation Innovation: A Case Study

of Carlink–A Smart Carsharing System

UCB-ITS-PRR-99-41

California PATH Research Report

Susan A. Shaheen

University of California, Davis

CALIFORNIA PARTNERS FOR ADVANCED TRANSIT AND HIGHWAYS


3/472

DYNAMICS IN BEHAVIORAL ADAPTATION

TO A TRANSPORTATION INNOVATION:

A CASE STUDY OF CARLINK—A SMART

CARSHARING SYSTEM

Prepared by

Susan A. Shaheen

October 1999


4/472

Copyright by

SUSAN ALISON SHAHEEN

1999


5/472-v-

Dynamics in Behavioral Adaptation to a Transportation Innovation:

A Case Study of CarLink—A Smart Carsharing System

Abstract

Most trips in U.S. metropolitan regions are drive-alone car trips, an expensive and

inefficient means of moving people. A more efficient system would allow drivers to share

cars. Such a system is often less convenient for travelers, but convenience can be

enhanced by deploying “smart” technologies in concert with shared-use vehicles and

transit.

The motivation for this research is to determine how the use of information and

communication technologies can enhance flexibility and mobility—and what value

travelers will place on these new transportation means. My dissertation, using new survey

research methods, examines CarLink, a smart carsharing service designed and deployed

under my direction. This dissertation integrates social marketing and learning theories

with human activity analysis approaches to explain the processes by which travelers can

and might accept a transportation innovation. I focus on methods of presentation and

learning to examine response dynamics. To explain the CarLink system to consumers, I

developed several informational media: a brochure, video, and “trial” clinic.


6/472-vi-

My dissertation is based on a longitudinal survey of responses to informational media

that I conducted with San Francisco Bay Area residents in the summer of 1998. The

survey results provide the attitudinal and belief data needed to evaluate dynamics in an

individual’s learning and valuing response to an innovation. To assist in evaluation and

interpretation, I also conducted four focus groups, which I moderated, in October 1998.

I found that willingness to use CarLink was influenced by the amount and type of

exposure, as predicted by social marketing and learning theories. Informational media

were used to teach targeted groups, and behavioral modeling (e.g., the video and drive

clinic) was introduced to develop participants’ confidence in adopting new behaviors. For

instance, participants who only read the brochure lost interest over time, while a large

majority of those who read the brochure, watched the video, and participated in the clinic,

stated that they would use CarLink. I documented the process by which individuals

moved through definable stages in the behavioral adoption model, from precontemplation

to contemplation, and in many cases into action.


7/472

-vii-

ACKNOWLEDGMENTS

I would like to thank the University of California Transportation Center, Partners for

Advanced Transit and Highways (PATH), the California Department of Transportation

(Caltrans), the National Science Foundation, the Dwight David Eisenhower

Transportation Fellowship Program, American Honda Motor Company, the Bay Area

Rapid Transit (BART) District, Lawrence Livermore National Laboratory (LLNL),

Teletrac, and INVERS for their generous contributions to the CarLink project. Each

helped make this dissertation research possible.

I also would like to thank DaimlerChrysler for their research donation, which allowed me

to collect much of the international carsharing data presented in Chapter 3. Thanks go to

Daniel Sperling and Conrad Wagner for their contributions to earlier versions of this

chapter, which they also co-authored. An earlier version of this chapter, “Carsharing in

Europe and North America: Past, Present, and Future,” was published in Transportation

Quarterly, Volume 52, Number 3 (Summer, 1998), pages 35-52. An updated version of

this paper, “Carsharing and New Mobility: An International Perspective,” has also been

submitted for publication in the Transportation Research Record (1999).

In addition, I would like to express my deep gratitude to Daniel Sperling (my chair and

advisor), Ryuichi Kitamura, and Richard Walters for reviewing my dissertation and in

supporting me in its design and implementation. Thanks also go to Kenneth Kurani and

Thomas Turrentine for their assistance with my research methodology and video


8/472

-viii-

development. Richard Katzev also provided valuable assistance in helping to manage the

CarLink field test, while I focused on my dissertation.

Furthermore, special credit goes to Montgomery Pfeifer and Creative Communication

Services (UC Davis) for their imaginative contributions to the CarLink brochure and

video production. Thanks also go to Daniel Sturges, who helped me with the brochure

design and naming “CarLink.”

I would also like to express my sincere appreciation to the CarLink project partners, who

supported the field test and my research, particularly: Victoria Nerenberg of BART;

Robert Uyeki of Honda R&D North Americas, Inc.; Clifford Loveland, Terry Parker and

William Tornay of Caltrans; Erma Paddack and Sal Ruiz of LLNL; Robert Tam of

PATH; Stanley Polk of Teletrac; and Uwe Latsch of INVERS. Special gratitude also goes

to the CarLink participants who brought my research to life.

Many UC-Davis students and researchers also deserve special credit for their assistance

with data collection and analysis, including: John Wright, Jie Lin, Terrence Polen,

Jennifer Ingersoll, Paul Miller, David Dick, Linda Novick, Bryan Jones, Robin Owen,

Monica Bally-Urban, and John McCann. Many undergraduate students also provided

valuable support with data entry, formatting, and proofreading, including: Safa Aliababi,

Jacqueline Au, Darshana Bhatt, Karen Goodwin, Gurpreet Kaur, Elaine Kim, Sydney

Nicholson, Joy Ogami, Amy Suvansin, Jennifer Tongson, and Anh-Kiet Tran.


9/472

-ix-

Additionally, I would like to thank Susan O’Bryant for her indispensable assistance in

managing the CarLink budgets.

Finally, I would like to express my deep appreciation to Timothy Lipman and my parents,

Joseph and Pauline Shaheen, who have provided me with love and encouragement in

achieving all of my goals.


10/472

-xi-

TABLE OF CONTENTS

ABSTRACT vACKNOWLEDGEMENTS vii

CHAPTER ONE: PROBLEM, OBJECTIVES, AND OVERVIEW

• SECTION 1.0: INTRODUCTION 1

• SECTION 1.1: DISSERTATION OBJECTIVES 6

• SECTION 1.2: DISSERTATION OVERVIEW 9

CHAPTER TWO: LITERATURE REVIEW


• SECTION 2.1: CARSHARING AND STATION CAR EXPERIMENTALRESEARCH 12

• SECTION 2.2: ATTITUDINAL STUDIES ON TRANSPORT POLICY 17

• SECTION 2.3: TRAVEL BEHAVIOR THEORY AND METHODS 19

• SECTION 2.4: SOCIAL LEARNING THEORY 23

• SECTION 2.5: SOCIAL MARKETING THEORY 32

• SECTION 2.6: INTEGRATION OF SOCIAL MARKETING, SOCIALLEARNING, AND ACTIVITY ANALYSIS THEORY

AND METHODS 39

CHAPTER THREE: CARSHARING AND NEW MOBILITY: AN

INTERNATIONAL PERSPECTIVE


• SECTION 3.1: HISTORY OF CARSHARING IN EUROPE 47

• SECTION 3.2: RECENT STUDY RESULTS FROM EUROPE 52

• SECTION 3.3: CARSHARING AND STATION CARS IN NORTH

AMERICA 53• SECTION 3.4: RECENT DEVELOPMENTS IN ASIA 59

• SECTION 3.5: INNOVATING THROUGH A CSO LIFECYCLE 61

• SECTION 3.6: USER CHARACTERISTICS AND MARKETPOTENTIAL 67

• SECTION 3.7: SOCIAL AND ENVIRONMENTAL BENEFITS OFCARSHARING 72

• SECTION 3.8: CONCLUSION 75


11/472

-xii-

CHAPTER FOUR: STUDY APPROACH

• SECTION 4.0: STUDY GOALS AND OBJECTIVES 81

•

SECTION 4.1: CARLINK PROJECT OVERVIEW 83• SECTION 4.2: OVERALL APPROACH AND HYPOTHESES 86

• SECTION 4.3: RESEARCH DESIGN 92

• SECTION 4.4: INDEPENDENT AND DEPENDENT VARIABLES 96

• SECTION 4.5: OPERATIONALIZATION OF KEY STUDY VARIABLES 102

• SECTION 4.6: DATA COLLECTION PROCESS 110

• SECTION 4.7: ANALYSIS TECHNIQUES 111

CHAPTER FIVE: DATA COLLECTION

•

SECTION 5.0: INTRODUCTION 115

• SECTION 5.1: SAMPLING FRAME AND PROCEDURES 117

• SECTION 5.2: SAMPLING BIAS 129

• SECTION 5.3: QUESTIONNAIRE DESIGN AND INCENTIVES 132

• SECTION 5.4: INFORMATIONAL MATERIALS 135

• SECTION 5.5: DRIVE CLINIC DEVELOPMENT 145

• SECTION 5.6: FOCUS GROUPS 150

CHAPTER SIX: BASELINE ANALYSIS OF STUDY POPULATION


• SECTION 6.1: DEMOGRAPHICS 155

• SECTION 6.2: TRIP CHARACTERISTICS AND MODE CHOICE 168

• SECTION 6.3: PSYCHOGRAPHIC CHARACTERISTICS 188

• SECTION 6.4: OTHER ISSUES 201

• SECTION 6.5: MODE CHOICE MODELS 206

CHAPTER SEVEN: CARLINK LONGITUDINAL SURVEY

RESULTS


• SECTION 7.1: PROJECT PARTICIPATION AND AWARENESS 215

• SECTION 7.2: RESPONSE TO CARLINK CONCEPT 219

• SECTION 7.3: CARLINK “USE”-A LONGITUDINAL PERSPECTIVE 226

• SECTION 7.4: CARLINK ATTITUDINAL RESPONSE 238


12/472

-xiii-

• SECTION 7.5: CARLINK USER MODEL 255

• SECTION 7.6: CARLINK FIELD TEST-WHO JOINED? 257

CHAPTER EIGHT: CARLINK DRIVE CLINIC SUMMARY


• SECTION 8.1: DRIVE CLINIC PURPOSE 266

• SECTION 8.2: DRIVE CLINIC METHODOLOGY 267

• SECITON 8.3: IN-VEHICLE QUERY: TOP TEN QUESTIONS ASKED 268

• SECTION 8.4: IN-VEHICLE QUERY SUMMARY 270

• SECTION 8.5: EXIT INTERVIEW SUMMARY 277

• SECTION 8.6: CARLINK CONCERNS 286


CHAPTER NINE: CARLINK FOCUS GROUP SUMMARY


• SECTION 9.1: CARSHARING LOT LOCATIONS 292

• SECTION 9.2: CARLINK LOT FEATURES 293

• SECTION 9.3: VEHICLE FEATURES 295

• SECTION 9.4: ECONOMICS AND BILLING 295


CHAPTER TEN: CONCLUSIONS


• SECTION 10.1: INFORMATIONAL MEDIA 304

• SECTION 10.2: FIRST HYPOTHESIS FINDINGS 316

• SECTION 10.3: TEST OF SOCIAL INTERACTION EFFECT 321

• SECTION 10.4: SECOND HYPOTHESIS FINDINGS 325

• SECTION 10.5: FOCUS GROUPS 327

• SECTION 10.6: EARLY TARGET ADOPTER PROFILE 329

•

SECTION 10.7: FUTURE RESEARCH 333• SECTION 10.8: STUDY CONTRIBUTIONS 334

APPENDIX I: CARLINK LONGITUDINAL QUESTIONNAIRES

• INITIAL QUESTIONNAIRE 340

• BROCHURE QUESTIONNAIRE 355


13/472

-xiv-

• 2 ND

PHASE EXPERIMENTAL (VIDEO) 369

• 2 ND

PHASE CONTROL QUESTIONNAIRE 379

• 3

RD PHASE EXPERIMENTAL (DRIVE CLINIC) 388

• 3

RD PHASE CONTROL 398

APPENDIX II: DRIVE CLINIC QUESTIONNAIRES

• IN-VEHICLE QUERY CHECKLIST 408

• EXIT INTERVIEW QUESTIONNAIRE 410

APPENDIX III: CARLINK FOCUS GROUP SUMMARIES

• FOCUS GROUP 1: CONTROL GROUP 414

•

FOCUS GROUP 2: PARTICIPANTS EITHER LIVING OR WORKING IN DUBLIN/PLEASANTON

REGION 429

• FOCUS GROUP 3: LAWRENCE LIVERMORE NATIONALLABORATORY PARTICIPANTS 435

• FOCUS GROUP 4: DUBLIN/PLEASANTON RESIDENTS 442

APPENDIX IV: PARTICIPANT EVENT PHOTOGRAPHS

•

DRIVE CLINIC PHOTOGRAPHS 448• FOCUS GROUP PHOTOGRAPHS 449


14/472

LIST OF TABLES• TABLE 3. 1: SUMMARY OF EXISTING NORTH AMERICAN CSOs 54

• TABLE 3.2: VEHICLE OWNERSHIP AFTER JOINING CSOs 73

• TABLE 3.3: CHANGE IN MODAL SPLIT 74

• TABLE 5. 1: NUMBER OF HOUSEHOLDS IN EACH GROUP 122

• TABLE 5.2: NUMBER OF INDIVIDUALS IN EACH GROUP 123

• TABLE 5.3: RESPONSES TO BROCHURE IMPROVEMENT QUESTION 140

• TABLE 5.4: RESPONSES TO VIDEO IMPROVEMENT QUESTION 144

• TABLE 5.5: RESPONSES TO DRIVE CLINIC IMPROVEMENT

QUESTION 148

• TABLE 6. 1: AGE 162

• TABLE 6.2: EDUCATION' 164

• TABLE 6.3: OCCUPATION 166

• TABLE 6.4: HOUSEHOLD COMMUTERS 172

• TABLE 6.5: CARPOOL TO WORK STATUS 173• TABLE 6.6: RENTAL CAR USER 176

• TABLE 6.7: RENTAL CAR USE: PURPOSE AND FREQUENCY

(EXPERIMENTAL GROUP) 176

• TABLE 6.8: RENTAL CAR USE: PURPOSE AND FREQUENCY

(CONTROL GROUP) 177

• TABLE 6.9: TRIPS IN HOUSEHOLD VEHICLE 178

• TABLE 6. 10: TRIPS WALKING, JOGGING, OR BICYCLING 178

• TABLE 6.11: TRIPS BY BUS 178

• TABLE 6.12: TRIPS BY BART/RAIL 178

• TABLE 6.13: TRIPS BY TELECOMMUTING 179• TABLE 6.14: OTHER TRIPS 180

• TABLE 6.15: NUMBER OF TRIPS MADE BY ACTIVITY

(EXPERIMENTAL GROUP) 180

• TABLE 6. 16: NUMBER OF TRIPS MADE BY ACTIVITY

(CONTROL GROUP) 181

• TABLE 6.17: MAJOR REASONS FOR CURRENT MODE 187

• TABLE 6.18: WHY NOT PARTICIPATE IN A SHARED RESOURCES

PROGRAM? 206

• TABLE 6.19: AUTO COMMUTER MODEL LOGISTIC REGRESSION

RESULTS 209• TABLE 6.20: TRANSIT COMMUTER MODEL LOGISTIC

REGRESSION RESULTS 211

-xv-


15/472

• TABLE 7. 1: WHY INDIVIDUALS AGREED TO PARTICIPATE 216

• TABLE 7.2: CARSHARING AWARENESS 217

• TABLE 7.3: NEW CARSHARING INFORMATION (SECOND PHASE) 218

• TABLE 7.4: NEW CARSHARING INFORMATION (THIRD PHASE) 219

• TABLE 7.5: THOUGHTS ABOUT CARLINK AFTER BROCHURE 220

• TABLE 7.6: THOUGHTS ABOUT CARLINK AFTER VIDEO 222

• TABLE 7.7: THOUGHTS ABOUT CARLINK AFTER DRIVE CLINIC 224

• TABLE 7.8: REASONS FOR NOT USING CARLINK 230

• TABLE 7.9: REASONS FOR USING CARLINK 232

• TABLE 7. 10: WHEN DID YOU REALIZE YOU MIGHT BE ABLE

TO USE THIS SERVICE? 235

• TABLE 7.11: CARLINK USER MODEL LOGISTIC REGRESSION

RESULTS 257

• TABLE 8. 1: CARLINK "DRIVE CLINIC" IN-VEHICLE QUERY

SUMMARY 271

• TABLE 8.2: PAYMENT METHOD 280

• TABLE 8.3: ACCESSORIES FOR CARLINK VEHICLES 285

• TABLE 10. 1: THOUGHTS ABOUT CARLINK AFTER BROCHURE 305

• TABLE 10.2: THOUGHTS ABOUT CARLINK AFTER VIDEO 308

• TABLE 10.3: THOUGHTS ABOUT CARLINK AFTER DRIVE CLINIC 312

• TABLE 10.4: REASONS FOR NOT USING CARLINK 320

• TABLE 10.5: REASONS FOR USING CARLINK 321

• TABLE 10.6: WHEN DID YOU REALIZE YOU MIGHT BE ABLE

TO USE THIS SERVICE? 323

-xvi-


16/472

-xvii-

LIST OF FIGURES

• FIGURE 3.1: REASONS TO PARTICIPATE IN CARSHARING 69

• FIGURE 5.1: DO YOU THINK THAT YOU WOULD USE CARLINK 129

• FIGURE 5.2: DO YOU THINK THAT YOU WOULD USE CARLINK

(NONRANDOM ONLY) 132

• FIGURE 6.1: COMMUNITY SIZE 156

• FIGURE 6.2: HOUSEHOLD SIZE 157

• FIGURE 6.3: INCOME 159

• FIGURE 6.4: AUTO OWNERSHIP 161

• FIGURE 6.5: MARITAL STATUS 163

• FIGURE 6.6a: EMPLOYMENT STATUS (EXPERIMENTAL GROUP) 165

•

FIGURE 6.6b: EMPLOYMENT STATUS (CONTROL GROUP) 165• FIGURE 6.7: VMT (VEHICLE MILES TRAVELED) 170

• FIGURE 6.8: COMMUTE MODES 174

• FIGURE 6.9: VEHICLE AVAILABILITY 175

• FIGURE 6.10: TOP REASONS INDIVIDUALS LIKE CURRENT MODES 185

• FIGURE 6.11: MODAL SATISFACTION 189

• FIGURE 6.12: VEHICLE HASSLE SCALE 191

• FIGURE 6.13: VEHICLE ENJOYMENT SCALE 193

• FIGURE 6.14: CONGESTION SCALE 195

• FIGURE 6.15: ENVIRONMENTAL CONCERN SCALE 197

•

FIGURE 6.16: EXPERIMENTER SCALE 199• FIGURE 6.17: HAVE YOU EVER BELONGED TO A TIMESHARE? 202

• FIGURE 6.18: HAVE YOU EVER BELONGED TO A HEALTH CLUBOR COUNTRY CLUB? 203

• FIGURE 6.19: HAVE YOU EVER BELONGED TO A FOODCOOPERATIVE? 204

• FIGURE 7.1: WOULD YOU USE CARLINK? 226

• FIGURE 7.2: CARLINK SATISFACTION (FIRST PHASE) 239

• FIGURE 7.3: CARLINK SATISFACTION (SECOND PHASE) 240

• FIGURE 7.4: CARLINK SATISFACTION (THIRD PHASE) 242

•

FIGURE 7.5: ALLOWS FOR SPONTINEITY (FIRST PHASE) 246• FIGURE 7.6: ALLOWS FOR SPONTINEITY (SECOND PHASE) 246

• FIGURE 7.7: ALLOWS FOR SPONTINEITY (THIRD PHASE) 247

• FIGURE 7.8: EMERGENCY RESPONSE (FIRST PHASE) 249

• FIGURE 7.9: EMERGENCY RESPONSE (SECOND PHASE) 249

• FIGURE 7.10: EMERGENCY RESPONSE (THIRD PHASE) 250


17/472

-xviii-

• FIGURE 7.11: CARLINK RELATIVE TO CURRENT MODES(FIRST PHASE) 251

• FIGURE 7.12: CARLINK RELATIVE TO CURRENT MODES

(SECOND PHASE) 252

•

FIGURE 7.13: CARLINK RELATIVE TO CURRENT MODES(THIRD PHASE) 253

• FIGURE 8.1: PAYMENT METHOD 282

• FIGURE 10.1: WOULD YOU USE CARLINK? 316


18/472

-xix-


19/472

-xx-


20/472

1

DYNAMICS IN BEHAVIORAL ADAPTATION

TO A TRANSPORTATION INNOVATION:

A CASE STUDY OF CARLINK—A SMARTCARSHARING SYSTEM

CHAPTER ONE: PROBLEM, OBJECTIVES, AND OVERVIEW

SECTION 1.0 INTRODUCTION

The vast majority of trips in U.S. metropolitan regions are drive-alone car trips. This

form of transportation is expensive and requires large amounts of land. As automobiles

gain market share, transit and ridesharing continue to lose market share. Today,

commuters are more likely to spend a longer time commuting than they did in the past

(Baldassare, 1991). Furthermore, attitudes toward commuting have become more

negative. Despite these trends, transit now accounts for less than two percent of

passenger travel, notwithstanding large subsidies (Vincent et al., 1994). A more efficient,

but often less convenient, alternative to private auto use would allow drivers to share

cars. By deploying “smart” transportation technologies in concert with alternative

vehicle-usage arrangements, the opportunity now exists to enhance transit services,

thereby improving their competitiveness with private, individually owned cars. At

present, several transportation providers are employing electronic and wireless

communication systems to facilitate the use and deployment of mobility services.


21/472

2

One of the problems motivating this research is the apparent inability of transit services

to satisfy the presumed high value placed on flexibility and mobility by urban and

suburban residents. The success of a transportation innovation depends in part on an

individual’s attitude toward the traditional auto (Cullaine, 1992). This dissertation, using

new survey research methods, examines one application of a smart transportation service:

shared-use vehicles (or “carsharing”). Since carsharing is being deployed throughout

Europe, Asia, and North America, it is important to develop an understanding of the

response to this emerging alternative in the U.S.

1.0.1 Smart Carsharing: Purpose and Goals

Through carsharing, individuals gain access to a shared fleet of vehicles for multiple uses

throughout a day without the costs and responsibilities of ownership. Instead of owning

one or more vehicles, a household accesses a fleet of vehicles on an as-needed basis.

Shared-use vehicles provide a shared community resource at transit stations (i.e., smart

station cars), neighborhoods, campuses, employment centers, resorts, etc. Travelers can

rent or lease a shared-use vehicle to drive to and from their homes, offices, shopping

centers, and transit stations. Carsharing can be thought of as organized short-term car

rental. Shared-use cars provide instant and convenient access to destinations that are not

conveniently accessible by transit.

The goal of carsharing is to help reduce traffic congestion, air pollution, and government

spending. Sharing vehicles could mean less traffic and fewer cars overall. Carsharing

could reduce congestion by cutting down on the number of vehicles needed by


22/472

3

households and society as a whole, and by facilitating and encouraging transit usage,

walking, and bicycling. For commuters especially, carsharing could offer an alternative to

getting to and from their destinations. Carsharing fleets could also be made up of ultra-

low-emission, energy-efficient cars. Because a carsharing organization would handle

maintenance and repairs, these duties would be completed properly and on schedule,

further reducing pollution and energy waste.

Carsharing could reduce government spending on arterial street systems and mass transit

by increasing transit ridership through added reverse commuters and midday, evening,

and weekend riders. Sharing vehicles might lessen the demand for parking spaces; by

serving multiple users each day, vehicles would spend less time parked. Moreover,

sharing could reduce the need for additional household vehicles to support a family’s

travel needs. Travelers could benefit by gaining the mobility of a car without having to

carry the full costs of ownership; transit operators could benefit by being able to tap a

much larger potential market; and society could benefit by diverting travelers from

single-occupant vehicles to transit for part of their trips.

Carsharing provides the potential to reduce the costs of vehicle travel to the individual as

well as society. When a person owns a car, much of the cost of owning and operating the

vehicle is fixed. The variable cost of using the owned vehicle is relatively low, and thus

the driver has an incentive to drive more than is economically rational. In contrast,

payments by carsharing participants are closely tied to actual vehicle usage. A carsharing

system in effect transforms the fixed cost of vehicle ownership into variable costs.


23/472

4

CarLink is the use of short-term rental vehicles and intelligent communication and

reservation technologies to facilitate shared-vehicle access at transit stations or other

activity centers for making local trips. Advances in smart system technologies have many

benefits for both public transportation agencies and private firms managing fleets.

Potential and existing users of these technologies range from smart paratransit to

carsharing organizations.

There are several smart technologies bundled into such a smart system. The central

technology is automatic vehicle location (AVL), which uses global positioning systems

(GPS) to pinpoint a position (up to the nearest meter). Digital Geographic Databases are

used with AVL to inform the driver/vehicle subscriber and the advanced traffic

management system (ATMS) of the vehicle’s location by street address (Casey and

Labell, 1996; Hardin et al., 1996). ATMS can employ state-of-the-art wireless

communications to connect the smart components and potential users through such media

as interactive kiosks and the Internet. Smart cards or keys, containing memory and a

microprocessor, allow customer access to a reserved vehicle and relay the billing and

reservation information to the vehicle and ATMS. Following my survey, a nine-month

CarLink field test was deployed with 12 compressed natural gas (CNG) Honda Civic

vehicles, which are linked to a smart system, to provide an intermodal transportation

service.


24/472

5

1.0.2 Research Approach

This dissertation focuses on the results of a longitudinal survey, conducted with San

Francisco Bay Area residents in the summer of 1998, which explored responses to the

smart carsharing concept over time. Furthermore, my study included a set of four focus

groups, which I moderated, with selected survey participants in October 1998.

As mentioned earlier, a field test of smart carsharing is being implemented through a

partnership of the Institute of Transportation Studies (University of California, Davis),

the Bay Area Rapid Transit (BART) District, American Honda, the California

Department of Transportation, Lawrence Livermore National Laboratory (LLNL),

Teletrac, and INVERS.

The carsharing model developed and explored in this dissertation is known as “CarLink.”

In the CarLink model, a fleet of vehicles is shared by three categories of participants:

Homeside Users, Workside Commuters, and Day Users. To facilitate the exchange of

vehicles and encourage transit ridership, BART serves as the principal access “hub.”

Homeside Users drive CarLink vehicles between home and the BART station daily and

keep the car overnight and on weekends for personal use. Workside Commuters take

BART to their workside station and drive a CarLink vehicle to and from BART and their

employment center.1 Day Users access CarLink vehicles at available “hubs” and use

them for tripmaking throughout the day.

1 In the CarLink field test, which launched in January 1999, the employment center is the Lawrence

Livermore National Laboratory, and the BART “hub” is the Dublin/Pleasanton station.


25/472

6

SECTION 1.1 DISSERTATION OBJECTIVES

Social learning and social marketing theories were used in this dissertation to explain the

processes by which travelers can and might accept or adapt to a transportation innovation.

“An innovation is an idea perceived as new by those who are confronted with it as an

option in choice…Reaction to an idea is quite different when one encounters it for the

first time, than when it has become routine” (Rogers, 1972, p. 86). I focus on methods of

presentation and learning to examine dynamics in target adopter response. Social learning

methods and the behavioral adoption model, developed by social marketing theorists,

were also tested.

To explain the CarLink system, I developed and examined several informational media,

including a brochure, video, and “trial” clinic. According to Magill et al. (1981), a

strategy should be established to accomplish the innovation communication (or diffusion)

objectives. In my study, communication objectives emphasized the disadvantages of

current modes, the advantages and disadvantages of smart carsharing, and how the

system works.

I also integrated the human activity analysis approach into the design of study

instruments, including the questionnaires, drive clinic, and focus groups. Activity

analysis examines the daily patterns of households and their members to capture and

explain travel behavior and choices. This methodology, integrating principles of

sociology, focuses on understanding behavior and lifestyle choices of study participants


26/472

7

and their households. Examples of the activity analysis approach in my study include

survey questions, such as how do you accomplish your weekly activities; how many trips

are taken by activity type per week; and how essential to your lifestyle is a household

vehicle.

For this dissertation, I directed a team of researchers in administering a quasi-longitudinal

survey over a four-month period. The longitudinal survey provides the attitudinal and

belief data needed to evaluate the social learning methodologies and social marketing

theory tested in my study. I use these data to assess dynamics in an individual’s learning

and valuing response to an innovation over time. Specifically, I test two “dynamic”

innovation response hypotheses.

• Hypothesis One: An individual's response to an innovation will be positively altered

by informational media (i.e., video, brochure, and drive clinic). Furthermore,

individuals who are not exposed to additional information about the innovation will

become increasingly negative toward it over time.

• Hypothesis Two: An individual’s valuing response to an innovation’s negative

mobility attributes (e.g., limitations on instant mobility) will become more positive

after learning more about the new technology. In contrast, an individual—unexposed

to additional information about the innovation—will respond the same to the negative

mobility attributes across the study (i.e., his or her response will remain unchanged).


27/472

8

These results are used to evaluate the validity of the social marketing framework as it

relates to the early phases of innovation adoption. Please see Chapter 2, Literature

Review, for a discussion of the theoretical and methodological areas relevant to this

dissertation.

Second, I evaluate the impact of social influence from friends, family, and colleagues

(i.e., during the contemplation phase of innovation adoption) on study participants’

response to the CarLink system. According to social marketing theory, social influence

plays a significant role in an individual’s decision to adopt a new product or approach.

Third, I assess the usefulness and effectiveness of three social learning methods in

explaining and demonstrating the CarLink system. The social learning methods tested

include written informational material, a modeling video, and an interactive drive clinic.

Furthermore, the drive clinic and longitudinal survey provide a practical test bed for

evaluating the “social desirability effect.” This is tendency of participants to overstate a

socially desirable position, especially in the presence of researchers.

Finally, I use the survey results to identify target audience characteristics of potential

CarLink adopters.


28/472


29/472

10

REFERENCES

Baldassare, M. (1991). “Transportation in Suburbia: Trends in Attitudes, Behaviors, and

Policy Preferences in Orange County, California.” Transportation 18: 207-222.

Casey, R.F. and L.N. Labell (1996). Advanced Public Transportation Systems

Deployment in the United States. Washington, DC, US Department of Transportation,Federal Transit Administration, Office of Mobility Innovation.

Cullaine, S. (1992). “Attitudes Towards the Car in the U.K.: Some Implications for Policies on Congestion and the Environment.” Transportation Research C 26A(4): 291-

301.

Hardin, J.A., R.G. Mathias and M.C. Pietrzyk (1996). Automatic Vehicle Location and

Paratransit Productivity. Tampa, FL, Center for Urban Transportation Research,University of South Florida.

Magill, K.P., E.M. Rogers and T. Shanks (1981). Improving the Diffusion of MassTransportation Innovations. Washington, DC, Urban Mass Transportation

Administration, U.S. Department of Transportation: 108 pages.

Rogers, E.M. (1972). Key Concepts and Models. Inducing Technological Change for

Economic Growth and Development. R. A. Solo and E. M. Rogers (ed.). East Lansing,MI, Michigan State University: 85-145.

Vincent, M.J., M.A. Keyes and M. Reed (1994). Urban Travel Patterns. Washington, DC,Office of Highway Information Management, Federal Highway Adminstration, US DOT:

152.


30/472

11

CHAPTER TWO: LITERATURE REVIEW

SECTION 2.0 INTRODUCTION

This dissertation applies social learning and social marketing to the field of transportation

innovation. My study approach also integrates social learning and social marketing

methods with activity analysis. Human activity analysis is a travel behavior methodology

focused on understanding the behavior and travel choices of households and their

members. The integration of these approaches is a synergistic one because all three are

interested in understanding behavioral dynamics (e.g., learning, technology adoption, and

travel behavior). By monitoring change, researchers can better understand how lifestyle

affects behavior and choices; how and why people might learn to change their behaviors;

and why individuals might adopt a new system or product.

This literature review covers several key areas relevant to this dissertation. The first is a

review of experimental research studies on carsharing and station cars (i.e., the empirical

focus of this thesis). The second section reviews relevant literature on attitudinal response

to a range of transportation policies. The third section discusses the role of travel

behavior theory— specifically activity analysis—and experimental situations in the

evaluation of a transportation innovation. The fourth and fifth sections address social

learning and social marketing theory, respectively.


31/472

12

SECTION 2.1 CARSHARING AND STATION CAR EXPERIMENTAL

RESEARCH

Very little recent experimental research has been published on the station car and

carsharing concepts. This section discusses results from published experimental studies

from Europe and the United States. A more detailed overview of international

developments is included in this dissertation in Chapter 3, Carsharing and New Mobility:

An International Perspective. The third chapter focuses mainly on actual programs and

their social and environmental impacts (e.g., increased transit ridership, reductions in

vehicle ownership) in contrast to controlled experimental research.

Carsharing organizations have conducted several studies in Europe. While most of the

surveys have small samples, employ simple questionnaires, and do not use control

groups, they do provide useful insights. Results from these studies are discussed in

Chapter 3. Below is a summary of published experimental studies that have been

prepared by researchers rather than by the implementing organizations themselves.

Recently, Steininger et al. (1996) published their survey results on a carsharing

organization in Austria. Using travel diaries and attitudinal surveys, they examined the

reasons why individuals became CSO members. They discovered that overall travel

declined after drivers became members. However, this study did not evaluate dynamics in

innovation response or the behavioral adoption process.


32/472

13

Massot et al. (1999) also published results from their demonstration of smart, electric

shared-use vehicles outside of Paris, France. In October 1997, Praxitele launched a full-

scale, 20-month demonstration of their advanced carsharing scheme. Praxitele, a 24-hour

self-service operation, was deployed in a Paris suburb (St.-Quentin-Yvelines) with 50

electric Renault CLIO cars. Using real-time data, researchers report that approximately

50 percent of drivers made only one trip, whereas regular users (i.e., those who have

made more than six trips since joining) comprised barely 10 percent of Praxitele’s

clientele, yet they made nearly half the trips. On average, 30 trips were made per

day

approximately 35 minutes in length. At the end of the project, approximately 600

trips were made per week. This program employed smart technologies to facilitate

vehicle usage and access. Results are based entirely on actual use rather than on travel

diaries or survey results.

In the United States, there have been two formal carsharing demonstration research

projects. The first was Mobility Enterprise, operated as a Purdue University research

program from 1983 to 1986 in West Lafayette, Indiana (Doherty et al., 1987; Muheim

and Partner, 1996). Each household leased a very small “mini” car for short local trips

and was given access to a shared fleet of “special purpose” vehicles (i.e., large sedans,

trucks, and recreational vehicles). Mobility Enterprise created a hypothetical cash flow

for its operations. They claimed economic viability, but only if the shared-use vehicle

services were run through an existing organization, such as a large fleet operator.


33/472

14

In the Mobility Enterprise field test, the mini vehicles leased to participants were used for

75 percent of the households’ vehicle miles of travel (VMT). In contrast, the shared-use

vehicle fleet was only used 35 percent of the time that it was available to households

throughout the experiment. (The Mobility Enterprise study findings did not provide the

percentage of a household’s total VMT that was made with special-purpose fleet

vehicles.) Although this program was considered a success in promoting shared use, this

service was discontinued because it was deployed as a research experiment.

Feinberget al.

(1986) conducted a survey with 83 undergraduate students of the Mobility

Enterprise concept, 46 of whom had previously participated in focus group interviews.

They found that a “…short description of the shared fleet concept produced the same

perceptions of success, feasibility, willingness to try, interest in joining, etc. as did

extended discussion of the concept [in focus groups]” (p. 16). Interestingly, this finding is

antithetical to this dissertation’s first hypothesis, i.e., an individual's response to an

innovation will be positively altered by increasing the number of informational media

(i.e., brochure, video, and drive clinic).

Furthermore, Feinberg et al. (1986) claim that the focus group interviews demonstrated

that financial savings would not define the success of Mobility Enterprise. Group

discussions demonstrated that potential consumers are not fully knowledgeable of full

vehicle ownership costs. Consequently, target adopters may not perceive a financial

advantage to Mobility Enterprise over private ownership. Feinberg et al . also argue that

“[b]y assuming a completely economically rational consumer, classical theory fails to


34/472


35/472

16

This project failed halfway through the planned three-year program. The primary

problem was the low and erratic income of many of the tenants. Many were later

discovered not to be credit-worthy for car ownership; many were students who shared an

apartment and were not actually listed on the lease. Another failing was the pricing

structure of STAR: it encouraged long-term, as well as short-term rentals. Long rentals

sometimes resulted in long-distance towing charges when the old, often poor-quality cars

broke down several hundred miles from San Francisco. STAR’s management tried to cut

costs by purchasing used, economy-class vehicles, but this resulted in high repair costs.

Also, STAR apparently offered too many models in each vehicle class, leaving members

dissatisfied when a particular car was unavailable (Russell, 1998).

One of the most recent demonstration projects in the U.S. was a two-year study of station

car rentals at Bay Area Rapid Transit (BART) District stations (Bernard and Collins,

1998). For this BART project, Cervero et al. (1994; 1996) conducted an early market

assessment of station cars using stated-preference survey methods. However, it is

questionable whether stated-preference methods accurately capture response to new

technologies (Kurani and Kitamura, 1996), with limited learning and time.

In 1999, two new demonstrations were launched, both involving smart carsharing. The

first is the CarLink field test, which started in January, and the second is Intellishare,

which launched in southern California in March. CarLink focuses on behavioral and


36/472

17

attitudinal response to shared use. In contrast, Intellishare examines user demand and

potential vehicle wait times at shared-use lots.

Chapter 3 of this dissertation provides a detailed overview of research and operations in

the areas of carsharing, station cars, and new mobility. New Mobility is a new

transportation approach that focuses on pairing clusters of smart technologies with

existing transportation options (e.g., rail, autos) to create a coordinated, intermodal

transportation system that could substitute for the traditional auto (Wagner and Shaheen,

1998; Salon et al.

, 1999).

SECTION 2.2 ATTITUDINAL STUDIES ON TRANSPORT POLICY

Recently, a few transportation attitudinal studies have been published of direct relevance

to this dissertation. A few of these papers examine the response of individuals to

transportation policies or alternatives. For instance, Baldassare (1991) conducted a study

in a southern California suburb (i.e., a rapidly growing industrialized region, similar to

the one in which this study was conducted) that examined whether or not traffic attitudes

affect commuting behavior and policy preferences (e.g., ridesharing). Baldassare found

that “[t]here is substantial opposition to transportation policies that involve financial or

lifestyle sacrifices, despite the experience of worsening commutes and the growing

perception of traffic problems” (Baldassare, 1991, p. 216). Further, transportation

attitudes did not appear to impact policy preferences. Those who perceived traffic as a

serious concern and rated freeways as unsatisfactory are no more in support of carpooling

than others who did not perceive transportation to be a problem.


37/472

18

In another study of attitudes toward alternative transportation and environmental policies

in the United Kingdom, Cullaine (1992) found the proposed policy success to be highly

dependent on autos and attitudes toward the environment and congestion. Cullaine found

that 69 percent of households that owned an auto thought that a vehicle was essential to

their lifestyle and did not want to sell it. While 19 percent stated that a car was not

essential to their lifestyle, they still did not want to sell their vehicle. Two percent

responded that their car was useful, but they had considered selling it. An additional 3%

thought that cars are useful, but they would consider selling theirs if public transportation

better suited their lifestyle. Finally, two percent of participants stated that it would not

take much to make them sell their car (Cullaine, 1992, p. 292). Furthermore, most

participants recognized that there are many traffic problems. For example, 85 percent of

respondents agreed that existing roads would not be able to handle increased traffic

projections. Seventy-nine percent also agreed that auto emissions were a major cause of

environmental problems. Thus, most participants seemed to recognize that autos cause

several problems.

Despite the recognition of these problems, the results showed that very few participants

were willing to adopt policies that restrict driving, particularly ones that impose pricing

penalities (Cullaine, 1992, p. 301). Indeed, the solution that received the greatest support

was improved bus and rail transportation. Interestingly, this dissertation offers

participants an alternative transportation system that could improve access to transit and

potentially reduce auto ownership, in contrast to pricing policies (e.g., road pricing).

Nevertheless, this dissertation’s methodology faces the same challenge of both these


38/472

19

studies, i.e., attitudes are not necessarily reflective of action. This dissertation builds on

these two studies by developing longitudinal survey methods (i.e., questions and scales)

for measuring individuals’ responses to an innovation relative to their current travel

modes. Each of the previous studies was based on a single-phase survey conducted in

1989, which did not examine how an individual’s attitudes toward a transport alternative

might change over time.

SECTION 2.3 TRAVEL BEHAVIOR THEORY AND METHODS

Travel behavior theory is relevant to this literature review because it offers a framework

for understanding transportation behavior and choices. In studying innovation response, it

is important to select a behavioral methodology that best captures individuals’ reactions

to the new concept as it becomes more familiar. In my study, I integrate the human

activity analysis approach with social learning and social marketing theories.

Traditionally, travel behavior methodology has been dominated by engineers and

economists who were motivated by the need to develop standard travel forecasting tools.

Indeed, the primary goal of the early travel demand models was to determine where

freeways should be built. There was little concern whether these models accurately

explained phenomena or were a reasonable representation of the underlying processes.

However, in the mid-1970s, a new phase in transportation demand modeling began to

develop—human activity analysis—integrating the insights of sociology. Jones et al.

(1990) define activity analysis as a framework in which travel patterns are analyzed at the


39/472

20

day or multi-day level. This type of analysis better reflects differences in behavior,

lifestyle choices, and activity patterns of the population.

In the activity analysis framework, travel is recognized as a derived demand, which is

based on an individual’s needs and desires to participate in activities that are spatially

separated (Pas, 1990). The focus of this approach is on understanding behavior rather

than on prediction. Typically, the household and its members are considered to be the

source of activity participation. Hence, the activity choices of household members are

mediated by a system of constraints, such as structure, resources, and relationships

(Kurani et al., 1996).

Kitamura (1988) states that the tools of more traditional travel-demand approaches—

especially, discrete-choice models and stated preference techniques—are fundamentally

different from those motivating the development of activity analysis. Garling et al. (1993)

argue that discrete-choice models cannot model interactions between individuals (e.g.,

within a household) or their choices. Further, the utility maximizing frameworks that are

central to discrete-choice models often reduce items into a single scale, even when some

items should not be combined (Kurani and Kitamura, 1996, p. 14). Consequently, other

approaches, such as activity analysis and social marketing, may provide more suitable

frameworks for understanding behavioral choices and change, particularly in response to

an innovation.


40/472

21

The most important theoretical and methodological trend in travel analysis over the last

20 years has been the development of activity analysis concepts and their increased

application. The federal Travel Model Improvement Program (TMIP) is premised on

bringing these concepts and methods into everyday, practical application. Activity

analysis inherently requires dynamic approaches to travel behavior—the study of

household activity and travel over time (Kitamura, 1988; Jones et al., 1990; Kurani and

Kitamura, 1996; Kurani and Lee-Gosselin, 1996). Furthermore, “Slovic et al. (1990)

argue that preferences are often constructed—not merely revealed—in responding to a

choice” (Kurani et al.

, 1996). This dissertation examines the longitudinal response to a

transportation innovation, based on several types and lengths of exposure to the CarLink

system. This is achieved by studying household response (and one to two individuals

from each household) over several months. At present, activity analysis is most often

studied for periods of just one day. Hence, behavioral adjustments and the methodologies

for studying these activity dynamics are weakly developed.

Many researchers are studying individuals’ reactions to unfamiliar, alternative

technologies. Indeed, they must rely on participant response to experimental situations in

order to understand the probable or even possible impacts they may have. This is

especially true of more innovative solutions (Kurani et al., 1996; Turrentine and Kurani,

1998). Kurani et al. (1996) claim that a cross-sectional study of current preferences is not

a sufficient method for assessing innovations. Several advanced survey designs have been

developed to help compensate for the limited understanding and experience that

participants have with a new technology (Golob and Gould, 1998). My dissertation


41/472

22

contributes to this body of knowledge by evaluating the impacts of informational

materials and an interactive clinic on participant response.

In evaluating a new technology, it is critical to document the processes of attitudinal

response, preference formation, and lifestyle evaluation relative to a household’s

exposure to an innovation. Kurani et al. (1996) have concluded that an experimental

situation will often elicit and engage the decision processes of its participants, often

revealing the participants’ lifestyle goals. These processes are often initiated by the

presence of a new technology. Since research into consumer responsiveness to

innovations (especially those embodying new values and performance attributes) must be

attentive to “processes” (Kurani et al., 1996), a longitudinal approach to evaluation is

used throughout my study.

As mentioned earlier, discrete-choice methods have often been applied to stated-

preference data in travel demand analysis. Kurani et al. (1996) have found that many

stated-preference studies of electric vehicle (EV) markets estimate huge price penalties

for limited-range vehicles (e.g., Beggs and Cardell, 1980; Morton et al., 1978). In

general, these studies rely on data from hypothetical-choice experiments in which

participants are presented with choice sets of the vehicle. Then, participants are asked to

identify the one vehicle, from each of the choice sets, they would be willing to purchase.

All vehicles are described by attributes that are common to all of the study vehicles, e.g.,

range. The attribute levels are varied over several trials to elicit different choices. With


42/472

23

these data, econometric models are run to estimate the partial utility values for consumer

preference of each attribute.

Kurani et al. (1996) argue that the underlying assumptions of consumer behavior in many

EV stated-preference studies are flawed. For example, these attitudinal studies assume

that the survey respondents have well-formed preferences for driving range. Second, they

assume that these preferences remain stable (or there must be enough longitudinal data)

to forecast changes in preferences. Third, oversimplified surveys often are designed to

encourage large sample sizes. Finally, these studies evaluate several vehicle attributes

that study participants have not yet experienced. For this dissertation, I designed an

attitudinal survey, which integrates social learning, social marketing, and activity analysis

approaches to address these concerns.

SECTION 2.4 SOCIAL LEARNING THEORY

Social learning theory emphasizes a continuous interaction among behavior, personal

factors, and environmental determinants. The relative influence of each factor is different

for various settings and behaviors. Social learning theory bridges the gap between

cognitively oriented rational decisionmaking models and behavioral theory. In this

framework, individuals are “…neither driven by inner forces nor buffeted by

environmental stimuli” (Bandura, 1977, p. 11). Rather, psychological processes are

explained in terms of a dynamic and continuous interaction of personal, behavioral, and

environmental factors. The environment can influence behavior by making it easier for

individuals to act. For instance, situational factors in the environment can influence


43/472

24

behavior (e.g., the close proximity of carsharing vehicles to a transit station could make it

easier for users to select this transportation option). A distinguishing feature of social

learning theory is that “symbolic, vicarious, and self-regulatory processes assume a

prominent role” (Bandura, 1977, p. 12). For instance, an individual might observe

another person’s behavior, reproduce it, and in replicating it, reinforce the modeled

behavior.

More traditional behavioral theorists have advocated a different learning framework.

From the behavioral perspective, learning can only occur after an individual performs an

activity and experiences its effects (i.e., trial-and-error learning) (Polley and Ven, 1996).

Cognitive theorists offer still another approach. They focus on rational processes and how

individuals’ preferences change as they undertake a course of action. For instance, once

an individual has decided to adopt an innovation they often reinforce this decision and, in

turn, become even more positive about this choice (Polley and Ven, 1996). Social

learning integrates these perspectives and advocates that “the capacity to learn by

observation enables people to acquire large, integrated patterns of behaviors without

having to form them gradually by tedious trial and error” (Bandura, 1977, p. 12).

Furthermore, social learning theory argues that as individuals gradually decide to adopt a

new behavior, they do not implement it instantly. “Among other effects, this slow

adaptation allows individuals to manage their anxiety in dealing with the newness of the

new behavior” (Andreasen, 1995, p. 268). This dissertation tests the validity of social


44/472

25

learning theory and the dynamics in the behavioral adoption process in response to a

social innovation (i.e., smart carsharing).

2.4.1 Use of Written Materials in Social Learning Theory

During the 1970s and 80s, social learning theory was applied to several social problems

(e.g., energy conservation and smoking). Perhaps most applicable is the experience of

energy conservation programs. According to Katzev and Johnson (1987), several types of

written informational materials, including brochures, posters, and labels, have been

developed to increase social learning to promote energy conservation measures. Many

studies have shown, however, that informational materials alone have little impact on

reduced energy consumption (e.g., Kohlenberg et al., 1976; Hayes and D.Cone, 1977;

Winett and Neale, 1979; Ester and Winett, 1981-1982; Anderson and Claxton, 1982;

Winett, 1986).

In contrast, one project claimed positive impacts resulting from an informational

campaign. Katzev and Johnson cite a two-year study of Heberlein and Baumgartner

(1985) in which researchers provided varying amounts of information to residential

consumers of a new electricity rate structure. This plan allowed consumers to reduce

energy costs by using electricity during off-peak times. Subjects in the “high” exposure

treatment had more accurate knowledge and favorable attitudes toward the rate structure

than subjects did in the “standard” treatment. Furthermore, individuals who received

more information consumed significantly less energy. Despite these results, it is difficult

to directly attribute this group’s behavior to the high level of media exposure. First, these


45/472

26

subjects received “feedback” from researchers (i.e., a letter congratulating them on their

efforts) and gifts for completing a test. Second, the impact of the rate change (e.g.,

savings to customers for using energy during off-peak times) was not isolated from the

information alone. That is, financial savings could also provide “feedback” to customers

that might affect behavior. Consequently, there is no way of determining the impact of

each informational stimuli on the subjects; therefore, more systematic research is needed.

This dissertation builds on the Heberlein and Baumgartner study and addresses several of

its weakness. It also looks at the impact of the three informational media on concept

response in both an experimental and control group. Economic feedback (e.g., savings

from implementing the innovation) are not included in this study’s design. Its greatest

weakness is that it does not examine actual use or behavior, but rather focuses on

attitudinal response to the concept.

Another example of an informational campaign is that of a smoking/anti-smoking

advertising experiment completed by Pechmann and Raneshwar (1993) in the early

1990s. This study focused on 304 seventh graders to determine whether or not cigarette

advertisements increase positive response towards smoking versus anti-smoking ads. In

this study, subjects were randomly assigned to a treatment (i.e., anti-smoking written ads)

or a control group (i.e., received ads unrelated to smoking). First, participants were

exposed to this material. Next, they were provided with information about a teenager’s

character and asked to evaluate this person on several attributes (e.g., personal appeal and


46/472

27

common sense). Individuals in the experimental group read a description about a teenager

who smokes. The control group read about a teenager who does not smoke.

According to Pechmann and Raneshwar, prior to the experiment, respondents did not

think that peers who smoked were any different than peers who did not (i.e., less

glamorous or mature). The participants did believe, however, that individuals who smoke

have less common sense and are less appealing than those who do not smoke. After the

experiment, researchers found that participants who reviewed the anti-smoking ads

produced more negative smoking inferences and judged the teenager smoker as having

less common sense and personal appeal. Furthermore, respondents who saw the anti-

smoking versus unrelated advertisements had a slightly higher tendency to discuss

negative smoker traits for a longer time than did the control group (Pechmann and

Ratneshwar, 1993). These results suggest that health-related education and anti-smoking

advertising can work in tandem. In conclusion, Pechmann and Raneshwar warn that anti-

smoking advertising may “wear out” over time. Therefore, funds should be allocated to

refresh these campaigns periodically. In the same manner, my dissertation builds on this

study’s results by looking at the effects of a carsharing brochure on an experimental and

control group over time. It also contrasts the response of experimental participants, who

receive more information over time, to the control group who only received the brochure.

Finally, it examines each of these groups’ comparative responses across the survey.


47/472

28

2.4.2 Use of Videos and Demonstrations in Social Learning Theory

As an alternative to written materials, Bandura (1977) emphasizes the significant impact

of observed behavior (i.e., videos or live demonstrations) on others. Furthermore, Ester

and Winett (1981-1982) have proposed the use of media-based “modeling” approaches to

enhance social learning. These approaches incorporate Everett Rogers’ diffusion of

innovation theory (Rogers and Shoemaker, 1971; Rogers, 1995). Rogers defines diffusion

as a special type of communication in which innovations are spread through members of

a social system (Rogers and Shoemaker, 1971). There are four key elements in the

diffusion process: innovation, channels of communication, time, and social system.

According to Rogers, there are five stages in the innovation adoption process, including:

knowledge, persuasion, decision, implementation, and confirmation (Rogers, 1995). This

theory models successive increases in the number of adopters over time. Diffusion

studies are concerned with the communication of new ideas from a source to a receiver.

In this framework, mass media channels are the most rapid and effective diffusion device

(e.g., television media). For Rogers, however, diffusion refers to an unplanned or

spontaneous communication, much of which is not applicable to this dissertation (Rogers,

1972).

In the area of energy conservation, Aroson and O’Leary (1982-1983) studied the impacts

of an individual demonstrating energy conservation in a college shower room. In a

demonstration, students watched a designated person turn off the shower while soaping.

When the designated person was present, there was an increase in student use of this

water conservation strategy (i.e., from six to 49 percent). This technique resulted in a


48/472

29

greater behavioral adoption rate than did written signs. Ester and Winett (1981-1982)

have also suggested the extension of approaches using “live” demonstrations. My

dissertation builds on this conclusion by developing a drive clinic or interactive trial

demonstration of participants with researchers. Rather than watching a researcher

demonstrate a new behavior (i.e., how to use the smart carsharing system), the participant

is taught how to use this system, which might reduce uncertainty and increase

understanding of the steps involved in using such an innovation.

Winett has also examined the impact of modeling on energy conservation behaviors. In

contrast to the previous examination, Winett’s studies involve a systematic replication of

field experiments, which build on the previous study. Winett (1986) argues that

presentation format and mode are critical in “modeling” impacts. Accordingly, videos

should employ individuals similar to those in the target audience, depict them in a range

of settings that display variations in the desired behavior, show constraints to the new

behavior, and provide an interesting story.

In his first study, Winett et al. (1982) exposed townhouse and apartment residents to a

20-minute video in which a married couple demonstrated energy-saving behavior and

others demonstrated wasteful use of energy. The scenes also emphasized the positive

consequences (e.g., money saved) resulting from energy conservation behaviors. A

second group viewed another video in which the same couple discussed energy problems,

without mentioning or demonstrating conservation strategies.


49/472

30

In the winter, individuals exposed to the modeling video consumed about 12 percent less

electricity during the five-week evaluation than those who viewed the second video and

17 percent less than the control group. These reductions were comparable to those

displayed by groups receiving either daily feedback (i.e., information about their daily

energy use) alone or feedback plus the modeling video. In the summer, subjects exposed

to the modeling video used about 12 percent less electricity during the four-week

evaluation period than they had previously. Subjects receiving feedback reduced

consumption by 19 percent, while those who received feedback and watched the

modeling video consumed 22 percent less electricity than they had previously.

In a second study, Winett (1983) replicated these findings, when the video was shown in

participants’ homes rather than during a group meeting as in the previous study. Energy

consumption decreased by approximately 13 percent during the winter evaluation period.

Furthermore, this effect was not increased by a second video exposure. Finally, the lack

of social interaction (i.e., the videos were not viewed with a group) was not a significant

factor in promoting energy conservation behavior.

In another study, Winett et al . (1985) extended these findings to a modeling video

broadcast over a cable TV channel. Participants, prompted to watch the program, reduced

electricity usage by approximately ten percent during the five-week evaluation and

generally maintained these reductions in a one month post-evaluation period. Video

exposure increased the group’s knowledge and adoption of energy conservation

behaviors depicted in the program.


50/472

31

Katzev and Thompson (1987) have concluded from Winett et al.’s extensive research that

carefully designed media interventions that model conservation behaviors may have

several effects:

1) leading individuals to save energy;

2) sometimes encouraging individuals to maintain the behavior

3) promoting the adoption of modeled behaviors; and

4) leading individuals to report a reduction in personal comfort due to behavioral

change.

Before accepting these conclusions, Katzev and Thompson caution that a more critical

review of study design is necessary. Indeed, it is important to rule out other explanations

before concluding that a stimulus accounted for a study’s results. Katzev and Thompson

argue that Winett et al.’s studies do not provide consistent control conditions in each

experimental study. Hence, it is difficult to compare results across the studies.

Furthermore, their results could have been confounded by the extensive recruitment and

follow-up methods used. For instance, throughout several studies, participants were often

visited by researchers who came to their homes to collect behavioral data. These visits

likely led to several participant-researcher conversations about energy-related matters.

Consequently, these contacts could have further sensitized participants to energy-related

issues, and in turn, promoted the observed energy reductions. Thus, it is difficult to

conclude that energy reductions reported in Winett et al.’s studies were due to the

modeling demonstrations alone and not the feedback of the researchers, for instance.


51/472

32

Despite extensive study in this area, much research still remains to be completed—

particularly in the area of careful study design. This dissertation contributes to social

learning theory by studying changes in response to an innovation due to several

informational media. A limitation of this dissertation is that it does not introduce

materials in varying orders to different experimental groups to assess the relative impact

of each stimulus. Rather, this dissertation looks at the impact of three instruments on an

experimental group over time. A further study is needed that examines the impact of

various devices, presented in different orders.

SECTION 2.5 SOCIAL MARKETING THEORY

Social marketing offers the second important framework relevant to this dissertation. It is

the application of concepts and techniques used in business to social behaviors. Social

marketing theory has been applied to health, family planning, child care, and the

environment (Kotler and Roberto, 1989; Andreasen, 1995). These techniques can also be

applied to transportation, as I have done in this dissertation.

Social marketing begins with targeted customers. It focuses on understanding a target

audience’s needs, wants, and perceptions and is directed at creating a “social” campaign

or product (e.g., anti-smoking campaigns and carsharing) (Andreasen, 1995).

“Social marketing recognizes that influencing behavior—especially behavior change—

cannot come about simply by promoting the benefits of some new course of action.


52/472

33

Careful attention must be paid to the nature of the behavior to be promoted (the product),

the ways in which it will be delivered (the place), and the costs that consumers perceive

they will have to pay to undertake it (the price)” (Andreasen, 1995).

Other key features of social marketing include an emphasis on program cost

effectiveness; the use of market research to design, pretest, and evaluate new programs;

careful market segmentation; and a recognition of competition (e.g., traditional auto

ownership and leasing are competition to carsharing).

Not surprisingly, social marketing builds upon other theoretical frameworks, including

traditional education, persuasion, social influence, behavior modification, and social

learning approaches by focusing on target adopters. Social marketing integrates and

improves upon those other approaches by addressing many of their weaknesses and

focusing on target adopters. Indeed, “[i]t often attempts to educate. It does seek to

motivate individuals to act. It does introduce group pressure when appropriate and it

often employs modeling and rewards to ensure the longer term success of its programs”

(Andreasen, 1995, p. 13).

Each of the building block frameworks for social marketing theory is reviewed below.

The traditional education approach emphasizes teaching and learning. Further, it assumes

that individuals will alter their behavior if they are educated on what needs to be done

and how to implement it. Andreasen (1995) points out several problems with the

educational approach. First, it assumes that if beliefs can be changed, then behavioral


53/472

34

change will result as well. Social marketing does not make this assumption; rather it

focuses on making a behavioral change occur and be sustained. Second, this approach

ignores the effects of social pressure. In contrast, social marketing recognizes that many

individuals engage in behaviors that they are not personally interested in or perhaps are

even opposed to (e.g., teenagers smoking due to peer pressure). Third, it assumes that

facts will have an intended impact. In many cases, however, a campaign may have

contrary effects. For instance, a breast cancer campaign, which emphasized that women

with family histories of cancer had a higher risk for this disease actually discouraged

women without cancer histories from conducting breast self-examinations.

The persuasion approach builds upon the educational framework. This model holds that

behavioral change will only occur when an individual is sufficiently motivated. The main

problem with this approach is convincing individuals to adopt this world view. In

contrast, social marketing promotes a user-centered approach to behavioral change,

which recognizes that a marketing campaign must begin with the customer’s perceptions,

needs, and wants (Andreasen, 1995).

In the social influence approach, public campaigns focus their attention on influencing

targeted community groups and collective behaviors. This framework addresses the cost

concerns of the behavior modification approach; yet, it has a few limitations. For

instance, this framework may be effective only in the following situations: 1) social

issues and norms of the targeted group are well understood; 2) pressures within the group

are influential; and 3) the behavior is socially important and visible (Andreasen, 1995).


54/472

35

An example of an effective social influence campaign might be an anti-smoking or anti-

drug campaign deployed in a secondary school.

The behavior modification framework is focused on two simple principles of learning

theory: first, individuals execute behaviors because they have been learned; and second,

these behaviors result in a positive outcome or reward. Until the early 1960s, a majority

of psychological learning theory assumed that individuals had to execute behaviors and

be rewarded to learn a new one. The main problem with this approach is that it is costly.

It typically must be implemented on an individual level rather than to a targeted group of

customers. “Social marketers recognize that, to have maximum social effectiveness in a

world of very limited budgets, one must focus on changing groups of consumers—not

individuals and not mass markets, but carefully selected segments” (Andreasen, 1995, p.

12).

In the early 1960s, Bandura and Walters (1963) contributed to learning theory when they

realized that children could learn new skills by simply watching other children. From this

finding, Bandura developed the social learning theory approach described earlier

(Bandura, 1977).

Social marketing builds upon and employs several social learning theory principles. For

instance, media (e.g., modeling videos and articles) can be used to stimulate learning by

targeted groups, and modeling can help develop an individual’s sense that they can


55/472

36

perform a new behavior. Nevertheless, the social marketing approach generally prefers

in-person training (e.g., drive clinics) over media devices, such as videos and brochures.

Similar to social learning theory, social marketing supports a gradual or dynamic

approach to behavioral adoption of a new product, concept, or service. Individuals move

through definable stages in adopting a new product (Maibach and Cotton, 1995). There

are four-stages in Andreasen’s social marketing behavioral adoption process: 1)

precontemplation, 2) contemplation, 3) action, and 4) maintenance.

Precontemplation is the first stage in the behavioral adoption process during which a

target population is introduced to the social product as a possible alternative to their

current behavior. The goal of this stage is to generate a

Dynamics in Behavioral Adaptation to a Transportation Innovation: A Case Study of Carlink–A Smart...

Documents

Transcript of Dynamics in Behavioral Adaptation to a Transportation Innovation: A Case Study of Carlink–A Smart...